Beta 299L - Shyvers Multiphone

[Whoop]

Groboto is good at producing abstract forms and shapes, but I have been interested to see if it can be used to model real world objects. The bumper car project was one.

This time around I decided to model a Shyvers Multiphone. 'What is one of these' I hear you ask.

The Shyvers Multiphone was a remote music selector system, used mainly in the Seattle and Tacoma area, between 1939 and the 1950s.

Unlike a conventional diner wall selector, you put your money in and spoke into the microphone at the top to a lady across town, who would play your selection back to you through the 4 inch speaker at the bottom, over special leased telephone lines.



The picture above right, of the selector box, is missing its base plate, so I added a shape for it, but I know it is wrong, as I have since found a better reference picture.

Modeling in Groboto was not too hard and I made heavy use of booleans. The part that gave me horrendous problems was meshing the top microphone grille. Because we don't have any parts that can create a blind hole, I had to make it in several sections and hope that Groboto could stitch them together. It went some of the way but made a nasty mess of some of the intersections.

I see beta M is available and L expired. I shall see if this fares any better.

Generally the booleans worked well. Sometimes they would mesh in groboto but would fail to export a .obj file without crashing. Sometimes they would export but not mesh in Groboto. I never did find any rhyme or reason behind the crashes.



As usual my workflow was to import into Silo, to manipulate the meshes some more and to create a UV map.

I then imported into Strata 3D for texturing and rendering. I was happy to see in Activity Monitor that Strata 3D maxed out, using all eight threads on the four cores of the i7 iMac. So far it is the only program I have seen that uses the second thread in these cores at all. So well done Strata. Rendering is very fast indeed and a joy.

[GroBoto]

Hey John,

Beautiful work. I'd not heard of this wonderful device -- ahead of it's time a hiFiPhone.

Truly clean meshes are a ways off as our first priority is a stable & robust meshing engine. If you could see inside the 'world of all possible topological configurations' that it must be able to resolve, you would know why it is taking time to track them down.

As for failed Meshes & such, Version L had several known bugs (more accurately, ones we learned of after release). Same is true of M -- but fewer.

I have (and do), confess to not keeping up on the documentation front. However, I always immediately post new release announcements here (Announcements Section), and the front page of the beta site. It will likely save you some pain if you snag the new builds when they go up (M has been there since the 20th -- N will post tomorrow or Wednesday.

Again, fine job -- great to see GroBoto tested in this sort of task.

Best Wishes,
Darrel

[Whoop]

I'd not heard of this wonderful device -- ahead of it's time a hiFiPhone.

I have seen quite a few of these devices here in the UK. I think someone bought a whole batch of them from the US some years ago. I remember seeing one for sale last October, at a jukebox show, which I took a photograph of:



As you can see, compared to the Wurlitzer remote selectors to its right and left, the Shyvers Multiphone was quite small.

[malachite]

Really nice work on the model. The render also looks good.

[Whoop]

Really nice work on the model. The render also looks good.

Kind words, thanks, I appreciate the response.